Document Type : letter to editor
Author
Augusta University, Medical College of Georgia, Augusta Georgia, USA 30912.
10.22038/jctm.2025.89590.1497
Abstract
OpenAI, a leading artificial intelligence research organisation based in California, is committed to developing technologies that align with human values and safety. Its latest flagship model, ChatGPT-4o, represents a significant advancement in natural language processing, featuring multimodal capabilities in text, voice, and vision. As part of the generative pre-trained transformer (GPT) family, ChatGPT-4o demonstrates enhanced speed, contextual understanding, and interactivity, making it a valuable asset in healthcare.
One emerging application is the development of intelligent virtual assistants for patient education and health management. ChatGPT-4o supports patients by providing real-time updates on medical conditions, treatment options, and device-related safety, including FAQs and guidance on remote monitoring and data security. Its use in assisting healthcare professionals with literature synthesis, diagnostics, and clinical decisions enhances medical efficiency and collaboration.
In cardiac care, machine learning (ML) techniques have shown remarkable sensitivity in interpreting electrocardiograms (ECGs), especially in detecting normal sinus rhythm. However, traditional ML models often struggle with complex arrhythmias due to artifacts and limited algorithm training. The implementation of unsupervised deep neural networks (DNNs), optimised for noise reduction and feature selection, has improved arrhythmia detection accuracy to 95%. These systems can identify subtle pathophysiological changes—such as fibrosis, hypertrophy, or chamber dilation—often undetectable to human observers, enabling earlier intervention for conditions like embolic stroke.
Implantable devices, including pacemakers and defibrillators, now utilise ML algorithms (e.g., random forests, convolutional networks) for continuous remote monitoring with higher predictive value than conventional scoring systems like CHA2DS2-VASc. While conventional pacemakers face challenges such as lead-related complications, recent innovations include biocompatible and antibacterial-coated leads, tissue-engineered materials, and leadless pacemakers, especially beneficial for dialysis patients. Looking forward, gene therapies to induce pacemaker activity in non-specialised myocytes and heartbeat-powered devices are under exploration.
While AI enhances personalisation, efficiency, and safety, it also raises concerns regarding data privacy, algorithmic bias, and ethical boundaries. Responsible deployment, interdisciplinary collaboration, and robust regulatory oversight are essential to maximising AI’s potential in shaping the future of cardiac care.
Keywords